Researcher, University of Luxemburg
1 paper at NeurIPS 2025
We propose a novel CAD reverse engineering method that mimics human design workflows by using multi-plane 2D cross-sections to reconstruct detailed, editable, and fully parametric CAD models.